Can someone explain mathematical induction to me?

-Whiplash-

New member
Joined
Jan 22, 2014
Messages
24
Yeah... I know it's sad, but I don't understand it, I'm doing math by correspondence and I don't have anyone to help me with it, I've watched up like 10 videos and at least read four guides but I still don't get it.

I've been in this calculus for 2 weeks now and I figured I skip it and coast off everything else, however, I got an assignment

for example I don't even know how to start with this question:

Use mathematical induction to prove that each formula is valid for all positive integral values of n.

-(1/2)-(1/4)-(1/8)-....-(1/2n)=1/2n-1 n∈N

I can do everything until the (k+1)=n part, I understand that K+1 is supposed represent the next part in the sequence algebraically but I get confused on how to show that and simplify.
 
Yeah... I know it's sad, but I don't understand it, I'm doing math by correspondence and I don't have anyone to help me with it, I've watched up like 10 videos and at least read four guides but I still don't get it.

I've been in this calculus for 2 weeks now and I figured I skip it and coast off everything else, however, I got an assignment

for example I don't even know how to start with this question:

Use mathematical induction to prove that each formula is valid for all positive integral values of n.

-(1/2)-(1/4)-(1/8)-....-(1/2n)=1/2n-1 n∈N

I can do everything until the (k+1)=n part, I understand that K+1 is supposed represent the next part in the sequence algebraically but I get confused on how to show that and simplify.
Let's take this in two parts.

First, the concept. There is something we want to prove, P, about all integers \(\displaystyle \ge\) j. So we first prove that P is true for j. Now, because it is true for j, that shows that there are one or more integers that are equal or larger than j for which P is true. With me so far? Now we choose an arbitrary one of those integers, k. But that means that P is true for k. This is all we know about k: it is an integer, P is true about it, and it is not less than j. We can use any or all those three facts in our proof. We now prove that P is true of k + 1, which is also an integer not less than j. We then say axiomatically P is true for all integers not less than j. Why does the axiom make sense? Let's say that j = 1. By our first step in the proof, we proved P true for 1. By our second step in the proof, we have shown that P is true for 1 + 1 = 2, but that means it is true for 2 + 1 = 3, which in turn makes it true for 3 + 1 = 4, and consequently true for 4 + 1 = 5, and so on without end. Is this concept clear in your mind?

Second, mechanics. With respect to proving the first step, this is usually not difficult, but it may require creativity. The second step is usually much tougher. There may be many ways to proceed, but one that I have frequently found helpful is this.

Let's say P is of the the form f(n) = g(n). Now P is true of k so f(k) = g(k).

Express f(k + 1) in terms of k. Express g(k + 1) in terms of k. They should be equal in terms of k. This is very abstract so I have an example. Look out for typos.

Prove \(\displaystyle If\ n\ \in \mathbb Z\ and\ n \ge 1,\ then\ \displaystyle \sum_{i=1}^n(2i-1)^2 = \dfrac{4n^3 - n}{3}.\) j is 1.

Step 1: \(\displaystyle \displaystyle \sum_{i=1}^1(2i-1)^2 = 1^2 = 1 = \dfrac{3}{3} = \dfrac{4 - 1}{3} = \dfrac{4 * 1^3 - 1}{3}.\) True of 1.

Step 2: \(\displaystyle \exists\ k\ such\ that\ k \in \mathbb Z,\ k \ge 1,\ and\ \displaystyle \sum_{i=1}^k(2i-1)^2 = \dfrac{4k^3 - k}{3}.\) With me to here?

Now I want to express \(\displaystyle \displaystyle \sum_{i=1}^{k+1}(2i-1)^2\ in\ terms\ of\ k.\)

That is not too hard because:

\(\displaystyle \displaystyle \sum_{i=1}^{k+1}(2i-1)^2 = \{2(k + 1) - 1\}^2 + \sum_{i=1}^k(2i-1)^2 = (2k + 1)^2 +\dfrac{4k^3 - k}{3} = 4k^2 + 4k + 1 + \dfrac{4k^3 - k}{3} = \dfrac{4k^3 + 12k^2 + 11k + 3}{3}.\)

Next I want to express \(\displaystyle \dfrac{2(k + 1)^3 - (k + 1)}{3}\ in\ terms\ of\ k.\)

This too is just algebra:

\(\displaystyle \dfrac{4(k + 1)^3 - (k + 1)}{3} = \dfrac{4(k^3 + 3k^2 + 3k + 1) - k - 1}{3} = \dfrac{4k^3 + 12k^2 + 11k + 3}{3}.\)

\(\displaystyle \displaystyle \sum_{i=1}^{k+1}(2i-1)^2 = \dfrac{4k^3 + 12k^2 + 11k + 3}{3} = \dfrac{4(k + 1)^3 - (k + 1)}{3} \implies\)

\(\displaystyle \displaystyle \sum_{i=1}^{k+1}(2i-1)^2 = \dfrac{4(k + 1)^3 - (k + 1)}{3} \implies\)

\(\displaystyle n \in \mathbb \ Z\ and\ n \ge 1 \implies \displaystyle \sum_{i=1}^n(2i-1)^2 = \dfrac{4n^3 - n}{3}.\ QED.\)

I do not guarantee that this is the most efficient method, but it works often enough for me.

So give your problem a try, and show us your work if you get stuck. It comes in time. You'll get it.
 
Last edited:
Greetings:

We wish to show that -1/21 - 1/22 - ... - 1/2n = 1/2n - 1 for all natural numbers, n (i.e., 1, 2, 3, ...).

To that end, we must show two things:

i) The claim holds for n = 1, and

ii) If the claim holds for some natural number k > or = 1, the it holds for k + 1.

From i) and ii), above, since it is true for n = 1, it must be true for n = 2. And since true for 2, it must be true for 3. Since true for 3, it is true for 4, and so on, ad infinitum. That is to say, the claim is valid for all natural numbers.

So, here we go.

i) Show true for n = 1: -1/21 = 1/21 - 1 and the claim holds for n =1.

ii) Suppose the claim holds for some natural number k. Then -1/21 - 1/22 - ... - 1/2k = 1/2k - 1. Now let us show that this (the induction hypothesis) implies the claim holds for k + 1. That is, we wish to show,

-1/21 - 1/22 - ... - 1/2k - 1/2k+1 = 1/2(k+1) - 1. But the first k terms on the left side sum to 1/2k - 1 according to the induction hypothesis and, therefore,

-1/21 - 1/22 - ... - 1/2k - 1/2(k+1)

= [-1/21 - 1/22 - ... - 1/2k] - 1/2(k+1)

= [1/2k - 1] - 1/2(k+1)

= 2/2(k+1) - 1/2(k+1) - 1

= 1/2(k+1) - 1. Hence the claim holds for k + 1.

Since i) the claim is true for n = 1, and ii) it is true for n = K + 1 if true for n = k, it follows by the principle of mathematical induction that such is true for all natural numbers, n.

I hope this helps.

Rich
 
Where did

Come from in your answer?
Whiplash

In my previous post, I gave you a method for attacking these problems and gave you an example of the method. My hope was that you would try to use that method on your own and SHOW us where you got stuck (if you did). You learn math by doing it. but maybe my example was too complex.

This is what you are intended to prove:

\(\displaystyle n \in \mathbb Z\ and\ n \ge 1 \implies \displaystyle \sum_{i = 1}^n\left(- \dfrac{1}{2^i}\right) = \dfrac{1}{2^n} - 1.\)

Your first step in a proof by (weak) mathematical induction is to prove that it is true for the smallest integer specified, in this case 1.

Can you do that?

Now I sometimes find it helpful to test the proposition for 2 and 3 to make sure that it seems to be true. This is not part of the proof, but it may help you "see" the general logic of a proof of the second step.

Once you have completed the proof of the first step, you KNOW that there is at least one integer for which what you are trying to prove is true. You hope there is more than one, but there is at least one. Choose any one of them and call it k. That gives you the starting point for the second step. This is frequently called the "inductive hypothesis," but it is not a hypothesis because you have proved the existence of at least one number that could be k. Anyway here is the starting point of the second step

\(\displaystyle \exists\ k\ such\ that\ \)\(\displaystyle k \in \mathbb Z,\ k \ge 1,\ and\ \)\(\displaystyle \displaystyle \sum_{i = 1}^k\left(- \dfrac{1}{2^i}\right) = \dfrac{1}{2^k} - 1.\)

In the proof of the second step, you may need to use the facts in red, and you will need to use the fact in blue. Notice that the first fact in red does not say that k = 1. Moreover, the facts in red automatically are true for k + 1 and k + 1 > 1. So you do have something to work with.

Now the method I suggest is to take the left hand expression for k + 1 and state it in terms of k. So

\(\displaystyle \displaystyle \sum_{i = 1}^{k + 1}\left(- \dfrac{1}{2^i}\right) = what?\)

Here is a hint

\(\displaystyle \displaystyle \sum_{i = 1}^{k + 1}\left(- \dfrac{1}{2^i}\right) = - \dfrac{1}{2} + \sum_{i = 2}^{k + 1}\left(- \dfrac{1}{2^i}\right) = - \dfrac{1}{2} + \sum_{m = 1}^k\left(- \dfrac{1}{2} * \dfrac{1}{2^m}\right) = - \dfrac{1}{2} + \dfrac{1}{2} * \left\{\sum_{i = 1}^k\left(- \dfrac{1}{2^i}\right)\right\} = what?\)

Now take the right hand expression and state it in terms of k. What do you get when you do that?

If the two expressions are the same, you have your proof.

Now PLEASE do the work yourself. Watching me do the work does not give you practice.
 
Whiplash:
Having simplified the sum to
[1/2k - 1] - 1/2(k+1), I wanted you to see where the final result, 1/2k+1 - 1 came from. In the interest of achieving a least common denominator, 2K+1, I multiplied top and bottom of 1/2k by 2, for 2/2k+1. Hence the expression becomes, 2/2k+1 - 1/2k+1 - 1 = 1/2k+1​ - 1 as desired.

Rich
 
Last edited:
Top